Cathy O'Neil: The era of blind faith in big data must end
Cathy O'Neil: Vessünk véget a big datába vetett vakhit korszakának!
Data skeptic Cathy O’Neil uncovers the dark secrets of big data, showing how our "objective" algorithms could in fact reinforce human bias. Full bio
Double-click the English transcript below to play the video.
the winners from the losers.
a győzteseket a vesztesektől.
that we don't understand
pontoznak minket,
and often hoping for.
by looking, figuring out.
közben figyeljük, számolgatunk.
what is associated with success.
mi minősül sikernek.
in written code.
leírt kód alakjában.
to make a meal for my family.
amikor főzök a családomnak.
of ramen noodles as food.
if my kids eat vegetables.
esznek zöldséget.
from if my youngest son were in charge.
más a siker mércéje.
he gets to eat lots of Nutella.
most people think of algorithms.
az algoritmusról gondolnak.
and true and scientific.
objektív, igaz és tudományos jószág.
a matekban, s félnek tőle.
blind faith in big data.
ha vakon hiszünk a big datában.
She's a high school principal in Brooklyn.
igazgató Brooklynban.
her teachers were being scored
modell"-lel pontozták,
what the formula is, show it to me.
mutasd meg nekem,
to get the formula,
azt mondta, hogy az matek,
told me it was math
a New York Post kérvényt nyújtott be,
a Freedom of Information Act request,
and all their scores
as an act of teacher-shaming.
the source code, through the same means,
szerezni a képletet, a forráskódot,
had access to that formula.
a képlet New Yorkban.
got involved, Gary Rubenstein.
ember, elkezdett vele foglalkozni.
from that New York Post data
talált 665 tanárt,
matekot is tanítottak.
for individual assessment.
ilyen sohasem történt volna.
with 205 other teachers,
recommendations from her principal
volt róla az igazgatójának,
of you guys are thinking,
és az MI-szakértők.
the AI experts here.
olyan következetlen algoritmust!"
an algorithm that inconsistent."
with good intentions.
lehet jó szándékból adódóan.
that's designed badly
silently wreaking havoc.
about sexual harassment.
szexuális zaklatás miatt.
to succeed at Fox News.
sikerre a Fox Newsnál.
but we've seen recently
to turn over another leaf?
hogy a helyzet javuljon?
their hiring process
21 years of applications to Fox News.
a Fox Newshoz beadott jelentkezési lapok.
stayed there for four years
to learn what led to success,
akik sikeresnek bizonyultak,
mely jelentkezők lettek sikeresek
historically led to success
a mostani jelentkezőkre,
to a current pool of applicants.
who were successful in the past.
nem válnak korrektté,
blindly apply algorithms.
if we had a perfect world,
don't have embarrassing lawsuits,
cégnek nincsenek kínos perei,
az adattudósoknak azt mondják,
ezért esetleg szexizmust
it means they could be codifying sexism
all neighborhoods
minden környék szegregált,
only to the minority neighborhoods
küldjük ki a rendőrséget
igen torzak lesznek.
we found the data scientists
where the next crime would occur?
a következő bűntett.
criminal would be?
ki lesz a következő bűnöző.
about how great and how accurate
but we do have severe segregations
de azért sok helyen
and justice system data.
és jogrendszeri adatokra.
the individual criminality,
recently looked into
during sentencing by judges.
was scored a 10 out of 10.
tízből 10 pontot kapott.
3 out of 10, low risk.
tízből három – kis kockázat.
for drug possession.
the higher score you are,
a longer sentence.
technologists hide ugly truths
a technikusok a csúnya igazságot
important and destructive,
and it's not a mistake.
building private algorithms
for teachers and the public police,
és a rendőrségnek szántakat is
the authority of the inscrutable.
hatalom gyakorlásából.
since all this stuff is private
will solve this problem.
sokat lehet keresni.
to be made in unfairness.
gazdaságilag racionális egyedek.
in ways that we wish we weren't,
még ha nem akarjuk is,
have consistently demonstrated this
of applications to jobs out,
alkalmas személy állásjelentkezését,
have white-sounding names
the results -- always.
ojtjuk be az algoritmusokba,
into the algorithms
milyen adatot gyűjtsünk,
about ramen noodles --
tudomást a zacskós levesről,
alapuló adatokban bízva
picking up on past practices
to emerge unscathed?
lesznek az algoritmusok?
we can check them for fairness.
the truth every time.
We can make them better.
algorithm I talked about,
kockázat algoritmusánál
we'd have to come to terms with the fact
hogy el kell fogadnunk a tényt,
arányban szívnak füvet az USA-ban,
smoke pot at the same rate
to be arrested --
gyakrabban tartóztatják le,
depending on the area.
négy-ötször gyakrabban.
in other crime categories,
más bűnügyi területen,
the definition of success,
kell a siker meghatározásával,
algorithm? We talked about it.
and is promoted once?
és egyszer léptették elő?
a cég kultúrájába.
that is supported by their culture.
the blind orchestra audition
are behind a sheet.
have decided what's important
distracted by that.
auditions started,
a vak zenekari meghallgatás,
went up by a factor of five.
száma ötszörösére nőtt.
for teachers would fail immediately.
szóló hozzáadottérték-modell.
the errors of every algorithm.
algoritmus esetleges hibáját.
and for whom does this model fail?
és kinél sikertelen a modell?
visszacsatolási hurkok
mérnökök gondoltak volna rá,
had considered that
only things that our friends had posted.
posztolta dolgokat mutatják nekünk.
egyik az itt ülő adattudósoknak szól.
one for the data scientists out there.
not be the arbiters of truth.
Nem lehetünk az igazság döntőbírái.
of ethical discussions that happen
folyó erkölcsi eszmecserét
for our algorithmic overlords.
nagyurai elszámoltathatók legyenek.
in big data must end.
vetett vakhit korszakának!
ABOUT THE SPEAKER
Cathy O'Neil - Mathematician, data scientistData skeptic Cathy O’Neil uncovers the dark secrets of big data, showing how our "objective" algorithms could in fact reinforce human bias.
Why you should listen
In 2008, as a hedge-fund quant, mathematician Cathy O’Neil saw firsthand how really really bad math could lead to financial disaster. Disillusioned, O’Neil became a data scientist and eventually joined Occupy Wall Street’s Alternative Banking Group.
With her popular blog mathbabe.org, O’Neil emerged as an investigative journalist. Her acclaimed book Weapons of Math Destruction details how opaque, black-box algorithms rely on biased historical data to do everything from sentence defendants to hire workers. In 2017, O’Neil founded consulting firm ORCAA to audit algorithms for racial, gender and economic inequality.
Cathy O'Neil | Speaker | TED.com